This was just an idea buzzing around in my head, but now I've got some time, I can ask about it.
Let's say I have a speaker, with a port tuned to 30Hz. The driver will unload <30Hz.
So what if I put a port tuned to, say 20Hz in, to stop some of the unloading, and reduce cone excursion around that frequency.
Would that work - anyone with any experience of this?
winISD won't let me use lots of different ports, so I decided I should ask. Maybe it's a good idea.
Let's say I have a speaker, with a port tuned to 30Hz. The driver will unload <30Hz.
So what if I put a port tuned to, say 20Hz in, to stop some of the unloading, and reduce cone excursion around that frequency.
Would that work - anyone with any experience of this?
winISD won't let me use lots of different ports, so I decided I should ask. Maybe it's a good idea.
That is not how a helmholtz resonator functions. Failed idea.
If you want to have two enclosure resonances look up the aperiodic bi-chamber enclosure.
If you want to have two enclosure resonances look up the aperiodic bi-chamber enclosure.
Earl Geddes had a bandpass box at RMAF in 2005 that worked like that I believe. I would contact him, or see if you can get his attention in this thread.
I know that Colin Whatmough of Whatmough Monitors/A1 Audio; used to tune one of his sub-woofers to two close but different frequencies, and now I can't find the old catalogue to check on the details.
What is the difference between two different sized pipes and a shelf port cut a 45 degree angle?? I tried that once and it seemed to improve performance.
What is the difference between two different sized pipes and a shelf port cut a 45 degree angle?? I tried that once and it seemed to improve performance.
I've done a little bit of rough testing on having 2 unequal length ports in an enclosure and they act like a port or ports that split the difference. Basically if you have a 16" port and an 10" port they will act like a pair of 13" ports at least as far as the tuning is concerned. Perhaps it could be useful for diminishing pipe resonances or something. I believe that Polk used to do this and called is ARC or something like that.
I've thought about having a slot port that tapers from a short length on one end to a long length on the other. Like 1" long that increases to 20" long on the other end within say 20" width. In theory I think that it would act like a 10.5" port for tuning purposes.
I've thought about having a slot port that tapers from a short length on one end to a long length on the other. Like 1" long that increases to 20" long on the other end within say 20" width. In theory I think that it would act like a 10.5" port for tuning purposes.
Josh Ricci said:
Josh,
Do you know how to use hornresp? One really cool feature, that's not well known, is that it can model vented and bandpass boxes.
Now that isn't a big deal, because a lot of programs can. The *cool* part is that horn response can model vents like you describe, where they're tapered. Horn response can also model vents that are flared at both ends.
It is the only program I am aware of (besides akabak) that can do this.
It's really interesting when you get into it, because you start to realize that you can get a huge bump in efficiency from a vented or a bandpass box if you use VERY large vents. The big drawback are resonances. But the original posters idea would deal with that.
I explored that to some extent in this thread.
If anyone needs me to clarify how you can model a vented/bandpass in hornresp, just let me know.
http://www.diyaudio.com/forums/showthread.php?threadid=128937
Josh Ricci said:
That's been my experience too, i.e a more broadband (lower Q) tuning which can be approximated via a single Karlson Coupler (pipe terminus slashed at an angle). This type of vent was patented some time ago as a new, unique breakthrough in venting, so probably legally strictly for personal use for now.
Can't remember the brand now, but IIRC it was either a small boutique vendor or start-up company, so may not even be around anymore as I don't recall reading about them since someone on one the forums asked about their performance claiams 10-12 yrs? ago.
GM
Josh Ricci said:
This is precisely what happens both theoretically and in practice. I typically use three, have used five, all different lengths. Just take the average if they are all the same diameter. (Thats not exact, but close enough).
unequal multiple ports
Hi there: If unequal multiple ports act "nearly" the same as a single port of the averaged size, why not use use the single average port? .... regards, Michael
Hi there: If unequal multiple ports act "nearly" the same as a single port of the averaged size, why not use use the single average port? .... regards, Michael
Re: unequal multiple ports
Because a long, large port has a severe resonance. And the larger and longer the port is, the lower in frequency and the higher in SPL the resonance is.
For instance, let's say you have a subwoofer that covers a range of 20hz to 80hz. Due to a port resonance, there's a peak at 160 hz. That *peak* can be more audible than the sub itself, particularly due to the Fletcher Munson curve.
It's not a small problem.
j.michael droke said:
Because a long, large port has a severe resonance. And the larger and longer the port is, the lower in frequency and the higher in SPL the resonance is.
For instance, let's say you have a subwoofer that covers a range of 20hz to 80hz. Due to a port resonance, there's a peak at 160 hz. That *peak* can be more audible than the sub itself, particularly due to the Fletcher Munson curve.
It's not a small problem.
Re: unequal multiple ports
The staggered ports act as one as a lumped mass, but have different resonances within the tubes when they have internal standing wave resonances.
j.michael droke said:
The staggered ports act as one as a lumped mass, but have different resonances within the tubes when they have internal standing wave resonances.
unequal multiple ports, "UMP"
Hi there, Earl and Patrick: Thanks for the explanations regarding unequal multiple ports. Is there a formula or design chart for utilizing UMP's?
As an example: 8cf box, single 5-inch port 13.3-inches long, box tuning Hz= 20hz, F3=20hz, Fs=19.7hz, if two UMP's are selected, would one be tuned below 20hz and the other above 20hz? How is this determined?
....regards, Michael
Hi there, Earl and Patrick: Thanks for the explanations regarding unequal multiple ports. Is there a formula or design chart for utilizing UMP's?
As an example: 8cf box, single 5-inch port 13.3-inches long, box tuning Hz= 20hz, F3=20hz, Fs=19.7hz, if two UMP's are selected, would one be tuned below 20hz and the other above 20hz? How is this determined?
....regards, Michael
find the correct length for one port and add and subtract some value, like an inch, to each, which need to be 1/2 the area of the single one. Its not complicated.
Sounds like it won't have the effect I expected.
Oh well.
Looks like an interesting thing to try to do, perhaps with an existing subwoofer...
All I need now is an existing subwoofer.
Oh well.
Looks like an interesting thing to try to do, perhaps with an existing subwoofer...
All I need now is an existing subwoofer.
gedlee said:
To calculate the effective port length take the equation along which paralleld inductances add:
given are L1, L2, L3 ... port lengths, port lenght correction included
Leff = 1 // ( 1/L1 + 1/ L2 + 1/L3 ... )
It's not exactly the average of L1, L2, L3 ...!
As You ask further paralleld port areas add like capacitors
Aeff = A1 + A2 + A3 + ...
The resonance frequency depends on the relation of area to length
Fr^2 ~ Aeff/Leff = (A1/L1 + A2/L2 + A3/L3 ...
Quite simple. One should keep in mind that narrower && longer ports resonate with a higher Q (on the pipe resonance) than one wider port alone. So there are diminishing returns with that tecnique. The benefit will strongly depend of the special situation.
hope this helps
ps: due to the square in Fr^2 ~ Aeff/Leff small errors in port length or area won't do to much. A deviation of 10% in Leff or Aeff will only cause Fr changeing about 5%. That is neglegible compared to atmospheric influences as air pressure, humidity etc.
The above correction to Mr. Geddes' note was done in case of someone building a double port of lets say 1" length + 10" length. The effective length will be much smaller than the average of both! It will be more around 1.1" instead of 5.5".
xpert said:
xpert
I said that the average was not exact, but when the deviations are a small part of the total lenght, as they should be, then the average is quite accurate.
And the statement about higher Q for narrower ports is not theoretically correct.
"And the statement about higher Q for narrower ports is not theoretically correct."
So why are they noiser than ports with a lower aspect ratio when used at high volume?
So why are they noiser than ports with a lower aspect ratio when used at high volume?
Something that lowers the Q of main port resonance is to place it near a wall or corner.
Lower Q produces less bass boost but reduces cone excursion (and thus distortion) over a wider frequency range, which was probably the initial goal...
Lower Q produces less bass boost but reduces cone excursion (and thus distortion) over a wider frequency range, which was probably the initial goal...
djk said:
Different effects altogether. The port noise is a nonlinear turbulence effect that depends on velocity which will be higher for a narrow port. The "Q" that is being takled about here is the Q of the first mode resonance of the port, NOT the Q of the box tuning. Again completely different things.
The simple math for the higher mode resonance of a tube does not contain the cross sectional area as a parameter. The damping of the higher modes (and hence the "Q") depends only on the boundary conditions, namely the radiation. Now there is a secondary effect on the radiation resistance with area, but for multiple tubes close together at the end, this will depend only on the total area and NOT of the subdivided areas, meaning that the "Q" of the subdivided ports will be the same as the single port - contrary to xperts claim.
So the staggered ports act as one as a lumped mass, as far as the box tuning goes.
Say we use four smaller ports vs one large one of the same area. The four smaller ports will have a circumference of twice that of the one large port. Is this a potential source of port compression at high volumes?
Say we use four smaller ports vs one large one of the same area. The four smaller ports will have a circumference of twice that of the one large port. Is this a potential source of port compression at high volumes?
- Status
- Not open for further replies.